How species interactions drive loss of fish biodiversity


This week’s featured research article has just been published in Current Biology by University of Adelaide researchers and examines how species interactions drive fish biodiversity loss in a high-CO2 world. The researchers studied species interactions in natural marine environments at underwater volcanic vents, where concentrations of CO2 match those predicted for oceans at the end of the century. They were compared with adjacent marine environments with current CO2 levels. This study was done in a shallow-water temperate kelp ecosystem using volcanic CO2 vents as natural laboratories to get a peek into what future ecosystems might look like. It further shows that forecasting effects of climate change on future ecosystems is impossible if we do not incorporate complex species interactions. The researchers showed in surveys and underwater experiments over three years, that in high CO2 marine environments one or two species of the smaller, behaviourally dominant fish proliferate while the less aggressive, and less common species disappear. The researchers state that if we look at the total number of fish we actually see that these increase under ocean acidification but local biodiversity is lost, and that there are increases in the abundance of food such as small crustaceans and snails. However, because the dominant species tend to win almost all combats with other species and are attracted to food much faster, their numbers rise. Small weedy species would normally be kept under control by their predators, but ocean acidification is also transforming ecosystems from kelp to low grassy turf, so habitat that protects these intermediate predators are being lost, and therefore losing these species. The researchers conclude that the result is a lot of what are known as weedy species – somewhat the marine equivalent to rats and cockroaches, plenty of them around but no-one really wants to eat them. A YouTube video about the research can be seen here  and the paper can be downloaded here.